BME280 Vs. DHT22 Temperature Sensor Shootout as Tested on the TTGO T-Display ESP32
Вставка
- Опубліковано 21 жов 2024
- Test Programme for this exercise is located here....
technocoma.blo...
After receiving multiple requests and information pertaining the the BMe280 sensor I have procured several for testing and comparison against the DHT22 sensor which I have used in previous videos. With a somewhat even price comparison as shown between amazon and Bangood between the two, power consumption for the DHT22 being higher in reading scenarios if power budget is a concern. The BME280 dos require some quick assembly if the pin headers are to be used.
Example links only below, do your own comparison!!!!
BME280 -
amzn.to/3yN34Xg
DHT22-
amzn.to/3Z2tI9i
Wiring the the TTGO is simple, one being I2C the other using a single GPIO. After checking, I'd modified one of the older DHT22 programs to accommodate both sensors for comparison. Initially i had ised a plastic bag to shield the sensors from the AC and fan but it appeared erratic over the course of 30 minutes, looking with the flir showed that the BME280 was too close to the TTGO. Adding extensions solved this, also using a second TTGO helped accuracy and putting all components into a project box further stabilized this. Variation between both BME280s was almost non-existent.
The second cold start test of 30 minutes revealed no skew in temperature, dispelling any internal heater possibility. Further tests of the pressure sensor showed almost no variation and tested close to the value on my Ipad.
In all tests it seems the BME280 did better than the DHT22.
#esp32 #arduino #maker #iot #maker
The reason your temperature was different is because the bme280 is chip temperature not ambient temperature, as I discovered by reading reviews on amazon that said just look at the datasheet, so whatever temperature it reports is likely to be a bit over ambient temperature, the integrated temperature sensor is only meant to improve the accuracy of the humidity and pressure sensors, if you need ambient temperature as well then I recommend DS18B20. That said I really like the bme280 for humidity and pressure :) Also now that I read this over, I am guessing its part of the reason you did the cold start over 30 minutes, so maybe its not really that bad even though its chip temperature.
Yes, these comments and complaints were the reason for the 30 minute tests. The ambient and enclosed case testing serves to show that in an encasement these work very well for temp measurements.
My BME280 was reading lower than a DHT11 though I don't have the best setups in the outdoor box I have them in right now. I put them up quickly this winter just for the fun of it outside on a portable solar cart setup I have in the field.
Thank you for your interesting multi-component comparisons and the TTGO T Display code and wiring info. I plan to replicate this and will try instead to compare another type of non-contact sensor the MLX90614 in place of the DHT-22 with this BME280. I have project in mind that needs to reliably trip at an adjustable threshold in a range of 60 to 70 degC before any thermal damage occurs. Its was very interesting to see the FLIR of the TTGO and breadboard.. as I touch mine it doesn't seem to get that warm but I rarely have it running as long as you did to assess these more accurately. Thanks Again. jimS
I put outside a DHT 22 , DHT 11, and BME280 and the BME280 temperature is off from the other two. I have ESPEasy running on ESP8266 so it was easy to put in a temperature calibration to bring it close to the two DHT sensors. I will have to do some more observation when it warms up again because the DHT 11 hit's it's bottom limit in the cold we are having now.
A very good video so I have joined your channel as your explanation and description was excellent thank you
Glad it was helpful
respect for the efforts. Good luck friend! very cool video and test. it became very very useful for me.
Glad it helped
Great video, thanks for the excellent analysis
Glad you liked it!
Excellent video, thanks! Boy, what a difference a year makes.......the cheapest for a SINGLE BME280 on Amazon now is $14 and three goes for around $30! That's some inflation!
all of these prices have exploded!!
Thanks for making this contribution. I found this interesting as I have spent time doing work similar to what you have done here, but primarily on the BME280. I'll offer my experiences; perhaps they will help you, or others.
1) These off-shore sources of BME280's are open to question. I have an Adafruit device just to check against the others; so far, no obvious differences in performance.
2) The BME280 WILL SELF HEAT if you run it too fast. The Adafruit library defaulted to this behavior (it has since been revised to stop this). Specifically, if left to default the BME280 will take readings at its maximum possible rate - way more than what we need for room temperature readings - and as a result the device heats up. The solution is to make it read only when you need it to. See the library documentation on "forced-mode". Once per second, or more, will eliminate the problem. Furthermore, even when setting the device to manual (forced) readings, if you call it in a loop continuously you will also self-heat it. Mine will rise some 2 or 3 degrees, F. I call for a reading every 5 seconds, which suits my purpose.
3) Another good way to compare any temperature sensors is to set a small fan blowing across both of them; more-or-less you get the "same air" on both sensors. They should be physically close together. Turn off the fan and watch the trend of each sensor; if one rises, then its self-heating.
4) As another mentioned, the DS18B20 is a great sensor, but temp only. Be warned, most all supply you see on auction sites are counterfeit. (see GitHub page by Chris Petrich, discover_fake_DS18B20.ino). Jameco has good prices on authentic parts. Counterfeit doesn't mean they don't work, but it just means you can't be sure of their compliance to specifications. I use both in various applications.
5) You mention averaging readings (at least I think you did...); the BME280 can do this for you, see the library for instructions on oversampling - the sensor can take care of this for you.
I like the convenience of the BME280, for sure. It offers a lot in one solution; but getting to trust the temperature readings has been a challenge. My desk clock displays temperature, standard pressure, relative humidity, and super accurate time using NTP - all on one 20x4 LCD display.
These TTGO modules look pretty cool - might have to give one a try.
Indeed, these were some of the perils mentioned in the technical analysis of the chip. I would agree in most cases where ambient air temp reading are taken, nobody would have a need to 30 reading per second and once every few seconds would be adequate.
Counterfeit chips are always an issue, especially on amazon. Considering the pricepoint there and bangood you can be sure that these offshore factories can be less than legit. Thats why the double blind tests are important.
is it possible to test any like dht22 with an oscilloscope?
Many new DSO scopes can decode digital values . Though I have also decoded it manually
Fresh study. :)
Thank you for not being Indian.
Why not enough whites on UA-cam ?
Then he must be a Chief.
@@silverback3633 If you want anything done right, go to a Chief. (Ask anyone in any Navy! - lol )
Having pride in something you have contributed nothing to is stupid.